Frozen section diagnosis of breast cancer.

In Savonlinna Central Hospital, Southeastern Finland, 371 frozen section examinations were made from September 1981 to December 1986 to detect breast cancer. There was one false positive diagnosis (0.7%) among the 147 lesions interpreted as carcinoma. Among the 224 biopsies, which were benign on frozen section investigation, four showed carcinoma in paraffin sections (1.8%). Thus, the diagnostic sensitivity of frozen section method was 97.3% (146/150) and the clinical diagnostic specificity 99.5% (220/221) when the diagnosis in the paraffin sections was used as reference. In the four false negative cases the tumour was small and limited to the breast without any evidence of metastases. Two of them were ductal carcinomas, one was microinvasive lobular carcinoma, and one intraductal non-invasive papillary carcinoma. The false positive case had benign intraductal papillomatosis. Our results suggest that the probability of a false diagnosis in frozen section examination increases with diminishing size of the lesion. We suggest that small lesions (less than 1 cm in diameter, or non-palpable) should not be subjected to frozen section examination to avoid unnecessary loss of neoplastic tissue during the preparation process. Instead surgically radical lumpectomy and careful investigation of paraffin-embedded tissue are recommended.